This calculator uses the fixed Stop Loss / Take Profit (pips) above — handy for a manual "what-if" check. The live Scanner & Auto-Trade page does not use these; it places its own stop using technical structure — see below.

When the scanner finds a signal, it does not use a fixed pip distance. It places the stop just outside the recent swing structure — the swing low for a BUY, the swing high for a SELL — so the stop sits where the trade idea is actually invalidated. The target looks at a wider window (roughly double the lookback) for a further structural level, rather than just the near edge of that same small range. If that technical reward isn't at least your minimum below, the trade is skipped rather than forced.

Applies automatically to every live Auto-Trade position — nothing to fill in. This mirrors the two-stage trail most systematic/CTA desks run: a hard jump to breakeven once the trade has banked a full 1R, then a volatility-adjusted (ATR-based) trail from there on — not a fixed fraction of distance, so a calm asset and a volatile one aren't trailed the same way.

Stage 1 — Breakeven jump: the stop stays untouched at its original level until the position's best price reached (peak progress) has moved a full 1R in its favor — i.e. as far as the original entry-to-SL distance, just in the profit direction. At that instant the stop jumps straight to entry: risk goes to exactly $0, no partial steps.

Stage 2 — Chandelier trail (ATR-based): past that point, the stop is recomputed every scan tick as best price reached − (ATR multiple × asset ATR), using each asset's own volatility (its ATR% of price, same figure driving position sizing on Portfolio) rather than a fixed fraction of the trade's own distance. The ATR multiple itself tightens as the trade runs further in profit — 3.0× ATR right after breakeven, narrowing down to 1.2× ATR as price nears the target — so the stop gives a freshly-profitable trade more room to breathe early on, then clamps down harder the deeper it's in the money. Once the trail pushes the stop past entry, "risk" goes negative: that's locked-in profit, and hitting that stop is booked as a win, not a loss.

The trail only ever tightens: it tracks the best price reached so far, so a temporary pullback doesn't loosen the stop back down. If price later reverses far enough to hit that trailed stop, the position closes for whatever is currently locked in (breakeven, or a partial profit) — not the original risk. If price instead reaches the target first, it closes for the full original reward. This runs every scan tick, so it needs scanning switched on to progress.

The Kelly Criterion (Kelly, 1956) computes the fraction of capital that maximizes long-run growth given a real edge: f* = W − (1−W)/R, where W is win rate and R is the reward:risk ratio from the calculator above. Betting more than full Kelly raises risk of ruin without adding long-run growth — most professional sizing uses a fraction of it.

VaR estimates the most you'd expect to lose over a horizon at a chosen confidence level, using the variance-covariance method (the same approach behind JPMorgan's RiskMetrics). CVaR / Expected Shortfall goes further — it's the average loss in the tail beyond VaR, so it doesn't understate risk in the worst-case scenarios the way VaR alone can.

—

Professional risk desks cut size after losing streaks and stop trading past a drawdown limit, instead of keeping risk constant regardless of recent results — this stops a bad run from compounding into a blown account.

Two different tools. The table below is a quick 3-point estimate (all wins / win-rate-weighted / all losses) for the single hypothetical trade set up in the Risk Calculator above — one asset, one win rate, one risk %, not the real book. The Monte Carlo simulation underneath is the real thing: it runs 1,000 correlated sequences of your actual Portfolio page — every active asset, its own win rate, allocation and risk %, correlated with the others via the same correlation matrix used in Portfolio → Diversification — to show the genuine range of outcomes and probability of ruin for the portfolio you've actually configured.

Run the simulation to see results.

This is different from the hypothetical simulator above — it reads the actual open positions from the Scanner & Auto-Trade page right now, and simulates how they could resolve together. Each open position is treated as a single win/lose outcome (it either reaches its TP or its SL), using that asset's rolling win rate. Because assets move together, outcomes aren't simulated independently: it uses a Gaussian copula — random outcomes drawn through the same rolling return-correlation matrix as the Portfolio page — so two highly-correlated positions (like EURUSD/GBPUSD) tend to win or lose together, the way they actually would in a real market move. This is a standard technique in quantitative portfolio risk (the same style of approach behind CreditMetrics-type correlated-default models), adapted here to open trades instead of credit defaults.

This only creates a proposal; nothing is applied automatically to the live engine.

Scans the assets you tick below on every closed candle, using your Live Chart strategy (MA cross + candlestick confirmation). When a signal fires on an asset with Auto-Trade on, an order is sent using the position size from Risk Management — nothing fires on assets left off.

Sized by Risk Management, and actually gated by it: the daily loss limit, the loss-streak/drawdown circuit breaker, Max Simultaneous Open Positions, and the Portfolio page's Heat Cap can all block a trade from firing — not just display a warning. A blocked signal is skipped and logged below like a rejected one.

Spread your trading capital across assets instead of risking it all on one. Each asset's slice becomes its own account balance for Risk Management sizing — this is capital allocation, not a live P&L view.

Expected Portfolio Profit is win-rate-weighted, not the best-case number above: for each open position it's winRate × reward − (1−winRate) × current risk (using each asset's own win rate), summed across the book — so it's what you'd actually expect on average, not what you'd get if every open trade won. Reward : Heat Budget compares total reward at stake to your full planned risk budget (Total Risk above) rather than just what's currently deployed, so — unlike Live Reward:Risk — it stays meaningful even before any trade has fired.

"Live" means it's computed from the actual open positions on Scanner & Auto-Trade, including any stop that's already been trailed up — as trailing frees risk (see Risk Management → Risk-Free Point), this ratio improves and Risk Free above grows, even with reward unchanged.

Combines each active asset's synthetic price history (same series used on the Live Chart page) weighted by its current allocation — a backtest-style illustration of how this mix would have moved, not real trade P&L.

High correlation between two heavily-allocated assets means they tend to win or lose together — that reduces real diversification even if capital looks split.